A mask for a face in motion
The Universal Voice mask supports vocal exercises and lets users sing words. This project focused on users aged 18–30.
The original sizes linked small to children, medium to women and large to men. These categories do not describe face movement during singing.
Measure the movement
The team reviewed mouth-opening data from Indian, Chinese, Polish and Irish populations. We then measured the nose-to-chin distance with the mouth closed and open in nine participants.
In this sample, the open-mouth distance had a strong relationship with face length. This result helped us choose the size measurements.
Explore the form, then design the mould
The team made early prototypes to study the outer shape, inner contact surface and outlet.
We could not print with flexible filament. We therefore designed a mould as another way to make the mask.
Choose six size groups
Six groups had the lowest misfit cost: 65.88. This digital sizing result still needs physical fit tests.
We compared face measurements from CEASAR (NL) for women and men aged 18–30. Face length and jaw width (bigonial breadth) were useful measurements. The analysis showed no strong relationship between them. The size system therefore uses both measurements.
From measurements to six size ranges
We processed DINED / CEASAR (NL) measurements and removed outliers. We then set size ranges and calculated each group centre (centroid).
Each size box defines a range. Its centroid gives a reference for the digital mask shape.
Six groups had the lowest misfit cost
The coverage target was 95%. Misfit cost is the total distance from points outside a size box to its nearest edge.
Check the Python settings and separate output
Check the Python settings and separate output
This screenshot sets six groups and a 95% target. The output shows four sizes, 90.70% coverage and a cost of 136.8098.
These results differ from the comparison above. We still need to check why they differ.
Rigid support, flexible contact
A rigid frame supports the mask. A flexible silicone surface contacts the face.
Nose and jaw anchors position the mask. Users can change the caps to adjust back pressure.
Use the size groups to define the mask shape
The team developed narrow, medium and wide versions in small and long sizes.
Measurement A defines the width. Measurement B defines the vertical length. These measurements connect the population ranges to six digital mask concepts.
More render views
More render views
The showcase carousel has the same six views. Expand this section to compare them.
Check six size groups and their limits
The team checked digital mask models for all six size groups.
We then examined four models at the measurement limits of Group 2, the largest group. These checks showed how the proposed mask met the limits of face length and jaw width.
Compare the mask with the jaw closed and open
The front and side views compare the digital mask with the jaw closed and open. They show how the face and mask shapes change before physical tests.